Advances and challenges of chimeric antigen receptor T cell therapy in digestive system malignancies

Abstract

Chimeric antigen receptor T cell therapy (CAR-T) has revolutionized the treatment of hematologic malignancies, but its success in solid tumors, particularly those of the digestive system, remains limited. Tumors of the gastrointestinal system, including gastric, colorectal, esophageal, hepatic, and pancreatic malignancies, represent a significant global health burden with high morbidity and mortality. Recent advances in antigen selection, chimeric antigen receptor design, delivery techniques, and combinatorial approaches have sparked renewed interest in CAR-T immunotherapy for these cancers. This article discusses recent progress in CAR-T development across the major digestive system tumors, outlines tumor-specific targets and clinical trials, highlights prevailing challenges and potential solutions, and proposes strategic directions for the next generation of CAR-T therapies in solid tumors.

Key Words: Chimeric antigen receptor T cell therapy; Digestive system tumors; Gastrointestinal cancers; Chimeric antigen receptor design; Immunotherapy

Core Tip: Chimeric antigen receptor T cell therapy (CAR) has achieved remarkable success in hematologic cancers but faces unique barriers in solid tumors of the digestive system. This article highlights recent progress in optimizing antigen selection, CAR engineering, and delivery strategies, while discussing tumor-specific targets and clinical trials in gastric, colorectal, esophageal, hepatic, and pancreatic cancers. By addressing challenges such as the tumor microenvironment and therapeutic resistance, we provide perspectives on advancing CAR T cell therapy immunotherapy toward safe, effective, and durable treatments for gastrointestinal malignancies.

INTRODUCTION

The ongoing clinical trials summarized in Table 1 highlight the translational progress of chimeric antigen receptor T cell therapy (CAR-T) in digestive tumors[1-5]. Notably, early findings from claudin 18 isoform 2 (CLDN18.2)-directed and glypican-3 (GPC3)-directed CAR-T trials demonstrate partial responses and disease stabilization in patients with advanced gastric and liver cancers, respectively[6-8]. Trials utilizing armored CAR-T constructs co-expressing interleukin-15 (IL-15) or dominant-negative transforming growth factor beta (TGF-β) receptors report improved persistence and immune activation in preclinical and early clinical settings[9]. Conversely, studies employing universal or ‘off-the-shelf’ CAR-T platforms are addressing manufacturing and accessibility bottlenecks[10-12]. Overall, these trials suggest a strategic shift from proof-of-concept safety assessments toward efficacy-driven optimization of CAR architecture and delivery modalities. This article focuses on recent progress in CAR-T therapy for gastric, colorectal, esophageal, hepatic, and pancreatic cancers.

Table 1 Current clinical trials of chimeric antigen receptor T cell therapy in the gastrointestinal system.

Interventions	Tumor type	Sponsor	NCT number
Claudin18.2 CAR-T	Digestive tumors	Shenzhen University General Hospital	NCT05620732
Cadherin 17 CAR-T	Digestive tumors	Chimeric Therapeutics	NCT06055439
Anti-mesothelin T naive/SCM hYP218 (TNhYP218) CAR-T	Digestive tumors	National Cancer Institute	NCT06885697
EpCAM CAR-T	Digestive tumors	Zhejiang University	NCT05028933
NKG2D/CLDN18.2 CAR-T	Digestive tumors	The Affiliated Hospital of the Chinese Academy of Military Medical Sciences	NCT05583201
Binary oncolytic adenovirus + HER2-specific autologous CAR VST	Digestive tumors	Baylor College of Medicine	NCT03740256
CDH17 CAR-T	Digestive tumors	Zhejiang University	NCT06937567
Claudin18.2 CAR-T	Digestive tumors	Suzhou Immunofoco Biotechnology Co., Ltd	NCT05472857
CLDN6 CAR-T	Digestive tumors	BioNTech Cell and Gene Therapies GmbH	NCT04503278
Claudin18.2 CAR-T	Digestive tumors	Legend Biotech USA Inc.	NCT05539430
CEA CAR-T	Digestive tumors	Changhai Hospital	NCT05240950
A logic-gated MSLN CAR-T therapy with a blocker receptor for HLA-A02	Digestive tumors	A2 Biotherapeutics Inc.	NCT06051695, NCT06682793
MSLN CAR-T	Digestive tumors	UTC Therapeutics Inc.	NCT06256055
MSLN CAR-T	Digestive tumors	CRISPR Therapeutics AG	NCT05795595
IL15 armored GPC3 CAR-T	Digestive tumors	Baylor College of Medicine	NCT05103631, NCT04377932
IL15 and IL21 armored GPC3 CAR-T	Digestive tumors	Baylor College of Medicine	NCT06198296, NCT04715191
GPC3 CAR-T	Digestive tumors	Second Affiliated Hospital of Guangzhou Medical University	NCT03198546
Universal CAR-T	Digestive tumors	Wondercel Biotech (Shenzhen)	NCT06653023
MSLN/GPC3/GUCY2C CAR-T	Digestive tumors	Second Affiliated Hospital of Guangzhou Medical University	NCT05779917
CD70 CAR-T	Digestive tumors	National Cancer Institute	NCT02830724
IL13R alpha2-specific hinge-optimized 4-1BB-co-stimulatory CAR	Digestive tumors	Jonsson Comprehensive Cancer Center	NCT04119024
CEA CAR-T	Digestive tumors	Chongqing Precision Biotech Co., Ltd	NCT06010862, NCT05415475, NCT06821048, NCT06126406, NCT06006390, NCT06043466
Irradiated PD-L1 CAR-NK cells plus pembrolizumab plus N-803	GC	National Cancer Institute	NCT04847466
Claudin18.2 CAR-T	GC	Peking University	NCT06353152
Claudin18.2 CAR-T	CRC	Second Affiliated Hospital, School of Medicine, Zhejiang University	NCT06946615
GUCY2C CAR-T	CRC	Beijing ImmunoChina Medical Science and Technology Co., Ltd.	NCT06718738
Universal CAR-T	CRC	Wondercel Biotech (ShenZhen)	NCT06653010
An armored GUCY2C targeting WD-01 CAR-T	CRC	Wondercel Biotech (ShenZhen)	NCT06675513
LGR5 CAR-T	CRC	Carina Biotech Limited	NCT05759728
Chemotherapy + allogeneic NKG2D CAR-T	CRC	Celyad Oncology SA	NCT03692429
MSLN CAR-T	EC	Memorial Sloan Kettering Cancer Center	NCT06623396
NKG2D CAR-NK cells	HCC	Zhejiang University	NCT07021534
GPC3 CAR-T	HCC	Shanghai Ming Ju Biotechnology Co., Ltd.	NCT06144385
GPC3 CAR-T	HCC	National Cancer Institute	NCT05003895
GPC3 CAR-T	HCC	Shenzhen University General Hospital	NCT05620706
GPC3 CAR-T	HCC	Renji Hospital	NCT05926726
B7H3 CAR-T	HCC	The Affiliated Hospital of Xuzhou Medical University	NCT05323201
GPC3 CAR-T	HCC	Zhejiang University	NCT06461624
IL-18 armored GPC3-CAR-T	HCC	Eutilex	NCT05783570
GPC3 CAR-T	HCC	CARsgen Therapeutics Co., Ltd.	NCT06560827
TGF-β receptor II armored GPC3 CAR-T	HCC	Zhejiang University	NCT05155189
GPC3 CAR-T	HCC	OriCell Therapeutics Co., Ltd.	NCT05652920
MSLN and claudin18.2 dual CAR-T	PC	Essen Biotech	NCT07066995
Claudin18.2 CAR-T	PC	CARsgen Therapeutics Co., Ltd.	NCT05911217
MSLN CAR-T	PC	Tianjin Medical University Cancer Institute and Hospital	NCT06760364
B7H3 CAR-T	PC	UNC Lineberger Comprehensive Cancer Center	NCT06158139

NCT: National Clinical Trial; CAR-T: Chimeric antigen receptor T cell therapy; SCM: Stem cell memory; EpCAM: Epithelial cell adhesion molecule; NKG2D: Natural killer group 2 member D; CLDN18.2: Claudin 18 isoform 2; HER2: Human epidermal growth factor receptor 2; CAR: Chimeric antigen receptor; VST: Virus-specific T-cell; CDH17: Cadherin 17; CLDN6: Claudin 6; CEA: Carcinoembryonic antigen; MSLN: Mesothelin; HLA: Human leukocyte antigen; IL: Interleukin; GPC3: Glypican-3; CD70: Cluster of differentiation 70; GUCY2C: Guanylyl cyclase C; IL13R alpha2: Interleukin 13 receptor alpha 2; 4-1BB: 4-1BB receptor; PD-L1: Programmed cell death ligand 1; NK: Natural killer; LGR5: Leucine-rich repeat containing G protein-coupled receptor 5; B7H3: B7 homolog 3; TGF-β: Transforming growth factor beta; GC: Gastric cancer; CRC: Colorectal cancer; EC: Esophageal cancer; HCC: Hepatocellular carcinoma; PC: Pancreatic cancer.

GASTRIC CANCER

Gastric cancer remains a leading cause of cancer death, particularly in East Asia[13]. Several tumor-associated antigens in gastric cancer, such as (human epidermal growth factor receptor 2) HER2, CLDN18.2, mucin 1 (MUC1), epithelial cell adhesion molecule (EpCAM), and mesothelin (MSLN), have been explored as CAR-T targets[14]. Notably, CLDN18.2-targeted CAR-T cells have shown encouraging responses in early clinical trials, with partial responses and disease stabilization observed in advanced gastric cancer (Table 1)[15]. Dual-targeting strategies (e.g., HER2 + CLDN18.2) are being pursued to address intratumoral antigen heterogeneity[16]. Despite these advances, antigen loss and poor T cell persistence remain barriers, which requires an in-depth exploration of potential solutions and ongoing research into overcoming these limitations. For instance, a tri-modular construct (CAR-T-epidermal growth factor receptor-IL13 receptor alpha 2-dominant-negative TGF-β receptor II) was developed for clinical application, which significantly elevated T cell proliferation and augmented functional responses[17]. This approach holds promise for overcoming resistance mechanisms commonly observed in suppressive tumor microenvironment (TME).

COLORECTAL CANCER

Colorectal cancer represents one of the most extensively studied digestive tumors for CAR-T applications[18]. Tumor-associated antigens like carcinoembryonic antigen, EpCAM, MUC1, guanylyl cyclase C, and CD133 have been targeted in numerous trials[19]. Recent studies have shown that combinations of CAR-T cells with vascular disrupting agents combretastatin A4 phosphate have shown enhanced intratumoral penetration[20]. Clinical trials targeting carcinoembryonic antigen and MUC1 are ongoing, with manageable safety profiles but limited efficacy to date (Table 1)[21,22]. Challenges such as antigen heterogeneity, immunosuppressive TME, and off-tumor effects are being addressed through logic-gated chimeric antigen receptors (CARs) and gene-edited designs incorporating checkpoint blockade or C-X-C chemokine receptor type 2 (CXCR2) enhancement to promote infiltration[23]. In addition, armored CAR-T cells engineered to express chemokine receptors or secrete proinflammatory cytokines can effectively counter the immunosuppressive TME and enhance CAR-T cell infiltration, thereby enhancing antitumor efficacy while minimizing systemic toxicity[24]. For example, studies have shown that CAR-T cells armored with CXCR4 infiltrate more into the TME and reverse its immunosuppressive characteristics, which significantly improves the efficacy[25].

ESOPHAGEAL CANCER

Esophageal squamous cell carcinoma and adenocarcinoma express several tumor-associated molecules, including HER2, EpCAM, and MSLN[26]. HER2-targeted CAR-T therapy has been trialed in HER2-overexpressing esophageal adenocarcinomas, with promising early findings[27,28]. However, high risks of off-tumor toxicity in HER2-expressing healthy tissues require careful antigen threshold modulation[29]. EpCAM-targeted CAR-T cells are being evaluated in mixed upper gastrointestinal tract tumors[30]. Strategies such as synthetic Notch receptors and local delivery (intra-tumoral or endoscopic injection) may mitigate systemic toxicities and enhance efficacy in esophageal cancer[31-33].

HEPATOCELLULAR CARCINOMA

Hepatocellular carcinoma is characterized by an immunosuppressive TME, making it particularly resistant to immunotherapies[34]. GPC3, and CD133 have been widely studied targets[35,36]. GPC3-targeted CAR-T cells have entered multiple clinical trials, demonstrating safety and transient tumor control[6,7]. Strategies to overcome the hypoxic and fibrotic microenvironment include co-administration of TGF-β inhibitors, gene-modified CAR-Ts resistant to exhaustion, and combinations with programmed cell death protein 1/programmed cell death ligand 1 blockade[23,26,37]. The dense stromal environment and poor CAR-T trafficking in hepatocellular carcinoma remain active research areas, with approaches such as IL-7/X-C motif chemokine ligand 1-engineered CAR-Ts showing improved efficacy in preclinical models[38].

PANCREATIC CANCER

Pancreatic ductal adenocarcinoma is notoriously resistant to all forms of immunotherapy[39]. CAR-T therapies targeting MSLN, and prostate stem cell antigen have been evaluated in early trials[40,41]. A study using MSLN-CAR-T cells demonstrated safety but limited objective response, likely due to immunologically “cold” TME and dense desmoplasia[42]. Incorporating oncolytic viruses or co-expressing chemokine receptors such as CXCR2 and C-C chemokine receptor type 2b may promote better trafficking[43,44]. Logic-gated CARs and armored constructs secreting IL-18 or IL-12 are under evaluation to boost local immune activation while minimizing systemic toxicity[45,46].

CHALLENGES AND POSSIBLE SOLUTIONS

The future of CAR-T therapy in digestive system tumors lies in overcoming the fundamental biological and logistical limitations that have thus far hindered its success in solid tumors. A major area of advancement is the development of multi-antigen targeting strategies[26]. Tumor antigen heterogeneity and antigen escape are leading causes of therapeutic failure[47]. Future CAR constructs will increasingly incorporate bispecific or tandem targeting domains that enable the recognition of multiple tumor-associated antigens simultaneously. Logic-gated CAR-T systems, can refine this by ensuring that CAR-T cells are only activated in the presence of specific antigenic combinations[46]. These strategies thereby enhance tumor specificity and minimize off-target cytotoxicity. Such approaches are especially promising in gastrointestinal cancers where tumor and normal tissue antigens often overlap.

Personalized medicine will also play a pivotal role in improving CAR-T outcomes. Advances in genomic, transcriptomic, and proteomic technologies are enabling a deeper understanding of the molecular and immunologic landscapes of tumors[48,49]. These insights will facilitate the identification of patient-specific neoantigens and glycosylation-specific epitopes that can be exploited for customized CAR design. Particularly in microsatellite-stable colorectal and pancreatic tumors, which are traditionally “cold” and lack immune infiltration, neoantigen-based targeting could help initiate a more robust immune response. For example, tumor-specific molecular profiling, such as whole-exome sequencing and RNA sequencing, can identify neoantigens that guide CAR design[48]. Additionally, advances in liquid biopsy and circulating tumor DNA technologies enable real-time monitoring of antigen evolution, allowing for adaptive CAR-T modifications[50,51]. In addition, personalized CAR-T manufacturing will likely be supported by artificial intelligence-assisted modeling of antigen expression patterns and immunogenicity to optimize targeting efficiency[52,53].

Another promising avenue involves modifying the TME to support CAR-T cell function. Solid tumors create an immunosuppressive milieu characterized by hypoxia, inhibitory cytokines, regulatory immune cells, and physical barriers such as desmoplastic stroma[54,55]. Future CAR-T constructs may include “armored” modifications, such as expression of proinflammatory cytokines like IL-12 or IL-18, or dominant-negative receptors that block immunosuppressive signals like TGF-β[39,56,57]. These engineered features can convert an immunologically “cold” tumor into a “hot” one, facilitating better immune infiltration and cytotoxicity. Additionally, CAR-T cells co-expressing chemokine receptors tailored to the TME may exhibit enhanced homing and persistence.

Beyond T cell engineering, delivery methods are undergoing critical refinement[58]. Regional administration routes such as hepatic artery infusion for liver metastases, peritoneal injection for gastrointestinal carcinomatosis, or endoscopic submucosal delivery for esophageal tumors may demonstrate promise in early trials[59]. These approaches improve CAR-T cell concentration at the tumor site while reducing systemic toxicity. Innovations in hydrogel-based delivery systems, nanoparticles, and scaffold technologies may also provide controllable release and spatial targeting of CAR-T cells, cytokines, or adjunctive agents within tumor regions[59-61].

Combination therapy represents a synergistic strategy to augment CAR-T efficacy. Checkpoint inhibitors, such as programmed cell death protein 1 and cytotoxic T-lymphocyte antigen-4 blockade, have shown potential in reactivating exhausted CAR-T cells and reversing TME-mediated immunosuppression[62]. Co-administration of oncolytic viruses can further enhance antigen release and dendritic cell activation, priming the tumor for CAR-T recognition[43]. In chemotherapy-refractory tumors, combining CAR-T cells with conventional cytotoxic regimens may debulk tumors, reduce antigen sink, and promote infiltration[43]. Preclinical and early-phase clinical studies are already demonstrating enhanced tumor control when CAR-T therapies are used in conjunction with these agents.

Finally, the scalability and accessibility of CAR-T therapy will be improved through the development of allogeneic, “off-the-shelf” cell products[63]. CAR-engineered natural killer (NK) cells and induced pluripotent stem cell-derived immune cells offer shorter manufacturing timelines and reduced cost while maintaining cytotoxic potential[64]. Their innate safety profile, particularly lower risks of cytokine release syndrome and neurotoxicity, makes them attractive for solid tumor applications. Several clinical trials are evaluating CAR-NK therapies targeting EpCAM, HER2, and MUC1 in digestive system cancers[65]. As gene-editing tools such as clustered regularly interspaced short palindromic repeats become more precise, the manufacturing of universal CAR-T and CAR-NK cell lines resistant to host immune rejection will become more feasible, paving the way for broader clinical deployment.

Collectively, these strategies represent a paradigm shift in how CAR-T therapy is conceptualized and applied in digestive tumors. The integration of multi-modal targeting, tumor-specific engineering, combinatorial regimens, and optimized delivery systems has the potential to finally unlock the full therapeutic potential of CAR-T in digestive malignancies. As these innovations converge, we embrace a future in which cell-based immunotherapies are not only effective in hematologic cancers but also transformative in the treatment of solid tumors.

CONCLUSION

While significant obstacles remain, the cumulative advancements in CAR-T engineering, delivery, and combinatorial approaches offer promising avenues for the treatment of digestive system tumors. Tumor-specific challenges, such as antigen heterogeneity, immune evasion, and physical barriers, are being addressed with increasingly sophisticated solutions. Future success will depend on the integration of genomic insights, novel antigen discovery, and well-designed combination regimens. As the field moves toward precision immunotherapy, CAR-T therapies may soon become integral components of multimodal treatment for gastrointestinal malignancies.